On-road transportation contributes 22% of the total CO2 emissions and more than 44% of oil consumption in the U.S. Technological advancements and use of alternative fuels are often suggested as ways to reduce these emissions. However, many parameters and relationships that determine the future characteristics of the light-duty vehicle fleet and how they change over time are inherently uncertain. Policy makers need to make decisions today given these uncertainties, to shape the future of light-duty vehicles. Decision makers thus need to know the impact of uncertainties on the outcome of their decisions and the associated risks. This paper explores a carefully constructed detailed pathway that results in a significant reduction in fuel use and GHG emissions in 2050. Inputs are assigned realistic uncertainty bounds, and the impact of uncertainty on this pathway is analyzed. A novel probabilistic fleet model is used here to quantify the uncertainties within advanced vehicle technology development, and life-cycle emissions of alternative fuels and renewable sources. Based on the results from this study, the expected fuel use is about 500 and 350 billion litres gasoline equivalent, with a standard deviation of about 40 and 80 billion litres in years 2030 and 2050 respectively. The expected CO2 emissions are about 1,360 and 840 Mt CO2 equivalent with a spread of about 130 and 260 Mt CO2 equivalent in 2030 and 2050 respectively. Major contributing factors in determining the future fuel consumption and emissions are also identified and include vehicle scrappage rate, annual growth of vehicle kilometres travelled in the near term, total vehicle sales, fuel consumption of naturally-aspirated engines, and percentage of gasoline displaced by cellulosic ethanol. This type of analysis allows policy makers to better understand the impact of their decisions and proposed policies given the technological and market uncertainties that we face today.

This content is only available via PDF.
You do not currently have access to this content.